Trajectory generation for controlling motion or other behavior of an autonomous vehicle may include alternately determining candidate actions and predicting future states based on the candidate actions. The technique may include determining a cost associated with the candidate action, which may include an estimate of a transition cost of the vehicle from a current or previous state to a next state. This cost estimate may be a lower bound cost or an upper bound cost, and the tree search may alternately apply only the lower bound cost or the upper bound cost or apply the lower bound cost or the upper bound cost according to a ratio or a varying ratio. The prediction of the future state may classify the dynamic object as a responsive object or a passive object based at least in part on a machine learning model, which may change the manner in which the dynamic object is modeled for prediction.
用于控制自主车辆的运动或其他行为的轨迹生成可以包括交替地确定候选动作并基于该候选动作来预测未来状态。该技术可以包括确定与候选动作相关联的成本,该成本可以包括对车辆从当前或先前状态到下一个状态的转变成本的估计。此成本估计可以是下界成本或上界成本,并且树搜索可以交替地仅应用下界成本或上界成本或根据比率或变化的比率应用下界成本或上界成本。对未来状态的预测可以至少部分地基于机器学习模型将动态对象分类为响应型对象或被动型对象,这可以改变动态对象被建模以进行预测的方式。
Vehicle trajectory control using tree search and/or proxy filtering with responsive proxy prediction
使用具有响应型代理预测的树搜索和/或代理滤波进行车辆轨迹控制
2024-03-29
Patent
Elektronische Ressource
Chinesisch
IPC: | B60W CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION , Gemeinsame Steuerung oder Regelung von Fahrzeug-Unteraggregaten verschiedenen Typs oder verschiedener Funktion |